Have you ever encountered a humanoid robot or a computer-generated face that looks almost real but still feels… off? That unsettling feeling is likely the uncanny valley in action.
Consider AVA, a “digital human” created by New Zealand tech startup Soul Machines for Autodesk. AVA expresses empathy through subtle cues like raised brows and head tilts, aiming to be virtually indistinguishable from a real person.
Yet, for many, AVA falls into a strange in-between space. She’s almost human, but not quite, creating an uneasy feeling. Like AVA, many realistic avatars, androids, and animated characters seem stuck in this disturbing zone – lifelike, yet unsettlingly wrong. This phenomenon is known as the uncanny valley.
Uncanny Valley: Definition
The uncanny valley describes our negative reaction to robots or other entities that appear almost, but not quite, human. The term was coined in 1970 by Masahiro Mori, a professor at the Tokyo Institute of Technology. Mori observed that as robots become more human-like, our affinity for them increases – but only to a point. Beyond that point, our response plummets into a feeling of unease, strangeness, and even revulsion.
The uncanny valley is the dip in comfort level that occurs when something is almost human, but not quite enough.
The Uncanny Valley Explained
In his essay for the Japanese journal Energy, Mori explained his observation. Our affinity for robots increases as they appear more human, until they reach a valley – the uncanny valley.
Mori uses the example of prosthetic hands to describe the uncanny valley: When we realize the hand, which at first site looked real, is in fact artificial, we experience an eerie sensation, together with its texture and coldness. When this happens, we lose our sense of affinity, and the hand becomes uncanny.
In an interview with IEEE Spectrum, Mori explained that his distaste for wax figures and early electronic prosthetic hands led him to the idea of the uncanny valley. These experiences triggered a sense of unease, prompting him to explore the concept of robots.
Uncanny Valley Examples
Several examples demonstrate the uncanny valley effect.
1. Telenoid
Telenoid, designed by Japanese roboticist Hiroshi Ishiguro, features a bald head, lifeless face, and lack of limbs.
2. Diego-san
Engineers at the University of California San Diego’s Machine Perception Lab developed the robot baby Diego-san to help parents better communicate with their infants. Even though the facial expression is sophisticated and intuitive in this infant robot, one expert says, along with a lack of detail in the eyes and forehead, can make the baby appear vacant and creepy,
3. Geminoid HI
Another of Ishiguro’s creations, Geminoid HI is an android replica. Ishiguro even used his own hair on the robot twin. Ishiguro says he created Geminoid HI to better understand what it means to be human.
4. Sophia
Designed by David Hanson of Hanson Robotics, Sophia displays a range of emotional expressions and is equipped with natural language processing capabilities.
5. Anthropomorphized felines
The 2019 live-action versions of The Lion King and the musical Cats brought the uncanny valley to the forefront of pop culture. The photorealistic computer animations of talking lions and singing cats that mimic human movements were just creepy to some fans.
Uncanny Valley: Science or Pseudoscience?
The validity of the uncanny valley as a scientific concept remains debated. Mori intended it as a guideline for robot designers, not a scientific concept.
Karl MacDorman, an associate professor of human-computer interaction at Indiana University, suggests Mori’s theory highlights the eeriness caused by mismatches in human likeness between appearance and touch, or appearance and motion.
How to Avoid the Uncanny Valley
Designers can reduce the uncanny valley effect by avoiding combinations of human and nonhuman, living and nonliving, or real and artificial features.
To move a robot or avatar beyond the valley, ensure that facial expressions match emotive tones of speech and that body movements are responsive to its emotional state. Special attention to the forehead, eyes, and mouth is also important.
Christoph Bartneck emphasizes matching a robot’s appearance with its abilities. He warns that the uncanny valley appears to be more of an uncanny cliff, finding that robots never recover once they fall into the valley.
Ultimately, understanding the uncanny valley allows creators to make conscious choices about the level of realism they aim for, and to navigate the fine line between fascination and revulsion in the quest to create human-like entities.
